272 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



and 0.7 mm. before shrinking). As the larvae 

 grow longer, their bodies grow increasingly deep 

 and they are therefore retained within the sampling 

 net. Evidence that larvae 3.0 mm. and longer 

 would be retained once they are in the plankton 

 net is given in table 28. 



Table 28. — Standard length and body depth at pectoral of 

 jack mackerel larvae 



[Measurements in millimeters] 



1 Data from Ahlstom and Ball (1964, p. 216-216). 



SAMPLING OF THE VERTICAL RANGE 



The vertical distribution of larvae was examined 

 for both the total range and quantitative diurnal 

 differences in abundance at different depths. The 

 data (table 29, after Ahlstrom, 1959: p. 119) 

 indicate that almost all the larvae were found 

 above 100 meters and that no larvae were found 



below 140 meters. Since net hauls are routinely 

 made from a 140-meters depth, the net passes 

 through the entire water column which can be 

 expected to contain jack mackerel larvae. 



The possibility of quantitative day-night differ- 

 ences (because of diurnal vertical migration) was 

 analyzed by examining the vertical distribution of 

 samples taken during the day and at night. These 

 ciaiparable day -night series were completed for 

 several stations. The 2-, 8-, and 19-meter strata 

 contain proportionally more larvae when sampled 

 in the day than at night. The data indicate that 

 there may be some slight tendency for the larvae 

 to migrate to the surface during the day ; however, 

 it is more likely that there was a change in the 

 water being sampled at station 94.80 (5403) 

 between the day and night series. Both sardine 

 and anchovy larvae are known to be distributed 

 somewhat deeper in the daytime (negative photo- 

 taxis) than at night (Ahlstrom, 1959), as are many 

 species of plankton organisms, so that a positive 

 phototaxis would need to be better documented 

 than is possible from our data. No definite 

 diurnal variation in vertical distribution is evi- 

 denced. 



Table 29. — Standard number of jack mackerel larvae taken in day and night hauls, by depths 



[Stations in parentheses; after Ahlstrom, 1959] 



•Region of the thermocline. 



VARIABLE DISTRIBUTION OF LARVAE 



The difficulties in assessing the variability of 

 these estimates have previously been discussed in 

 the section on sources of error in sampling eggs. 

 Since the proper statistical model is not known at 

 this time, no good estimate of error can be given. 

 It was shown in the case of eggs that for a large 



number of interpolations in space and time the 

 error tended toward zero. The error of the esti- 

 mates of abundance for the larvae is probably as 

 great as the error of the estimates for egg abund- 

 ance, and very likely it is even greater. This may 

 be particularly true for the estimates of abundance 

 of the older size categories which are based on 

 fewer than 10 observations. 



